Introduction

In matters of compression, the SIC concentrates on methods preserving as much as possible the image quantitative properties (radiance) whereas classical methods rather aim at preserving the visual quality of the compressed images. This lead the SIC to develop methods based on Wavelet transforms, both lossy and lossless.

The applications reside mainly in satellite image compression, where the SIC is responsible for several studies and operational projects for the ESA and for Eumetsat.

Outline of the projects...

Image compression module for the Ground segment of Meteosat Second Generation.
The SIC acts as compression expert developping the compression modules of MSG's ground segment, for VCS, Eumetsat's main contractor for the ground segment of MSG.

Image compression techniques for near real-time data dissemination for Meteosat Second Generation
Due to the huge volume of data produced, MSG data need to be compressed before being send to the end-users. The study consisted in building a database of simulated MSG image and optimizing compression parameters for that database.

Meteosat four seasons
In order to illustrate the usefulness of Wavelet-based compression methods, the SIC compressed one year of Meteosat data, i.e. more than 32000 images so they would fit on a double CD-ROM.

Meteosat image compression using wavelet transform
Aiming at compressing Meteosat images, the SIC developped high-quality wavelet-based compression methods for the ESA. In particular, an automatic optimization of the compression parameters was developped.

Use of lossy image compression schemes for data distribution
Various compression methods were compared, in view of using them for near-real time dissemination of satellite image data (MSG). Particular attention was paid to progressive compression and the possibility of performing selective decompression.